[patch V3 00/13] entry, x86, kvm: Generic entry/exit functionality for host and guest

From: Thomas Gleixner <tglx@linutronix.de>
To: LKML <linux-kernel@vger.kernel.org>
Cc: x86@kernel.org, linux-arch@vger.kernel.org,
	Will Deacon <will@kernel.org>, Arnd Bergmann <arnd@arndb.de>,
	Mark Rutland <mark.rutland@arm.com>,
	Kees Cook <keescook@chromium.org>,
	Keno Fischer <keno@juliacomputing.com>,
	Paolo Bonzini <pbonzini@redhat.com>,
	kvm@vger.kernel.org
Subject: [patch V3 00/13] entry, x86, kvm: Generic entry/exit functionality for host and guest
Date: Thu, 16 Jul 2020 20:22:08 +0200	[thread overview]
Message-ID: <20200716182208.180916541@linutronix.de> (raw)

This is the 3rd version of preparing KVM to handle TIF_NOTIFY_RESUME before
exiting to guest mode.

The first attempt is available here:

    https://lore.kernel.org/r/20190801143250.370326052@linutronix.de

The second version which tried to provide generic infrastructure for KVM
can be found here:

    https://lore.kernel.org/r/20191023122705.198339581@linutronix.de

This started the whole discussion about entry code correctness which led to
the complete overhaul of the x86 entry code, non-instrumentable sections
and the other goodies which are in 5.8.

The remaining bits are rather straight forward:

  1) Move the entry/exit handling from x86 into generic code so other
     architectures can reuse the already proven to be correct code instead
     of duplicating the whole maze and creating slightly different beasts
     again.

     This utilizes the noinstr sections. The correctness of the confinement
     of non-instrumentable code can be verified with objtool on
     architectures which support it. The code code itself is clean.

  2) Provide generic work handling functionality for KVM

  3) Convert x86 over to use the generic entry/exit code

  4) Convert x86/KVM to use the generic work handling functionality.

#4 finally paves the way to move POSIX CPU timer heavy lifting out of
interrupt context into task work. Several hundred patches after the initial
attempt which was small and simple :)

The patches depend on:

    git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git x86/entry

The lot is also available from git:

    git://git.kernel.org/pub/scm/linux/kernel/git/tglx/devel.git x86/entry

The posix timer patches will be posted seperately as they are mostly
independent (except for the KVM task work detail).

The stub functions for the architecture specific parts which need to be
provided by architectures utilizing this are completely documented in the
corresponding header file which explains the diffstat.

Thanks,

	tglx

----
 arch/Kconfig                        |    3 
 arch/x86/Kconfig                    |    1 
 arch/x86/entry/common.c             |  632 ++----------------------------------
 arch/x86/entry/entry_32.S           |    2 
 arch/x86/entry/entry_64.S           |    2 
 arch/x86/include/asm/entry-common.h |  135 +++++++
 arch/x86/include/asm/idtentry.h     |   39 --
 arch/x86/include/asm/ptrace.h       |   15 
 arch/x86/include/asm/signal.h       |    1 
 arch/x86/include/asm/thread_info.h  |    5 
 arch/x86/kernel/cpu/mce/core.c      |    4 
 arch/x86/kernel/kvm.c               |    6 
 arch/x86/kernel/signal.c            |    2 
 arch/x86/kernel/traps.c             |   24 -
 arch/x86/kvm/Kconfig                |    1 
 arch/x86/kvm/vmx/vmx.c              |   11 
 arch/x86/kvm/x86.c                  |   15 
 arch/x86/mm/fault.c                 |    6 
 include/linux/entry-common.h        |  398 ++++++++++++++++++++++
 include/linux/entry-kvm.h           |   80 ++++
 include/linux/kvm_host.h            |    8 
 kernel/Makefile                     |    1 
 kernel/entry/Makefile               |    4 
 kernel/entry/common.c               |  364 ++++++++++++++++++++
 kernel/entry/kvm.c                  |   51 ++
 virt/kvm/Kconfig                    |    3 
 26 files changed, 1150 insertions(+), 663 deletions(-)